fleutot+openmoko at gmail.com
Wed Jun 6 12:57:50 CEST 2007
On 6/6/07, Casper van Donderen <casper_de_dondergod at hotmail.com> wrote:
> Can you still edit that main panel [...] because the color of choice at
> the moment is orange [...]
I think that when we talk about UI, we don't primarily think of skins and
colors: these rather fall in the "bells & whistle" category. What makes a UI
good or bad are rather questions like (examples):
- does it take many clicks to register an unknown caller's number in the
phonebook? Can I easily extract phone numbers from SMS bodies?
- is the menu organized in a sensible way, or do I need to know that in
order to *stop* the system I have to click on the "start" menu button? (a
favorite of mine on windows pocket: if you click OK after having types a
SMS/e-mail, it cancels it instead of sending it).
- there's a limited space for shortcuts. Is the phone configured so that
default shortcuts actually correspond to the most common actions? How far
can I customize it, if the default shortcuts don't fit my usage scheme?
- screen estate is also a valuable asset. Is it attributed wisely, with more
space devoted to widgets I'm most likely to use, or is there a pointless
whirling logo that takes half of the screen and forbids me from reading an
- if I've typed an e-mail, changed my mind and decide to send it as an SMS,
how easy is it? Conversely, if 99% of what I send are SMS, and only 1%
e-mail, I won't be happy if I need to answer 5 modal dialogs asking me
questions which don't make sens for an SMS. Generally speaking, modal
dialogs are extremely obtrusive, and should only be used when something
- I don't mind alpha blending effects and stuff like that, *provided that it
doesn't break responsiveness*. If it takes 3 seconds to open a text composer
due to an impressive animated logo, by the 3rd time I'll use it I won't
think "woot, neat animation!" anymore, but rather "stop wasting my time and
CPU cycles, and bring me my fucking editor RIGHT NOW!"
- If I'm in front of a building requiring a entrance code, and I know I've
got that code somewhere in my phone, but I'm not quite sure about the
reception date, nor whether it was a mail, an SMS or a note I typed myself,
how tedious will be the search? Will I have to put up with some braindead
interface inspired by "clippy the MS-Office trombone"? Will there be enough
search criterions? Conversely, will I have to fill a 3 pages form before
every trivial search request?
- will I need to spend 1/2h configuring a PC to talk with my phone, or will
it simply be seen as a USB thumbdrive?
Look at Steve Jobs' first presentation, it was straight-to-the-point. What
iPhone users will love are stuff like:
- easy, reliable, unobtrusive unlocking mechanism
- every relevant feature can be reached in 3 or 4 clicks, and these click
sequences are easy to remember because they follow common sense.
- no need to navigate through 12 menus to rotate your navigator, so that you
can read a page not designed for mobile devices (users of winCE know what I
- you won't accidentally press touches with your cheek when phoning, thanks
to the cheek-clicking protection mechanism.
- it's fast and easy to retrieve a contact number, even in real-life
conditions (e.g. on a bicycle, with direct sunlight on the screen, and no
way to hold a stylus)
All of these stuff aren't impressive because they require l33t coding
skillz: they're impressive because they correctly anticipate the way real
users will use the phone in their real lives. That's much more important,
and much more difficult to get right, than 3D glowing and spinning widgets.
And I think openmoko can lead to real improvements in this domain, if:
- the UI is easy to modify, so that experiments aren't too expensive (I'm
talking about global reorganizations here, not about color schemes tuning)
- there's an efficient feedback system, typically a set of wiki pages
dedicated to ergonomics, field usage reports, polls, etc.
- some people with the right social skills make the UI improvement effort
These are stuff which only open source can do correctly, which is excellent
news for openmoko.
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the community